运用阶次跟踪和奇异谱降噪诊断齿轮早期故障  

Early Fault Diagnosis of Gear by Order Tracking and Singular Spectrum Denoising

作　　者：康海英[1] 祁彦洁[2] 阎文[1] 田燕[3] 郑海起[1] 

机构地区：[1]军械工程学院火炮工程系,石家庄050003 [2]徐州空军学院基础部,徐州221000 [3]军械工程学院装备指挥与管理系,石家庄050003

出　　处：《振动．测试与诊断》2010年第6期662-664,共3页Journal of Vibration,Measurement & Diagnosis

基　　金：国家自然科学基金资助项目(编号:50375157;50775219)

摘　　要：针对齿轮箱升降速过程中振动信号非平稳的特点,将阶次跟踪分析与奇异谱降噪技术相结合,提出了一种针对齿轮早期故障的诊断方法。首先对齿轮箱加速时测得的瞬态信号进行时域采样,再对时域信号进行等角域重采样,转化为角域伪稳态信号;然后对角域信号进行奇异谱降噪处理,以减小背景噪声的影响;最后对降噪后的信号进行阶次谱分析。通过对齿轮箱早期故障信号的分析表明,该方法能准确地识别出齿轮的故障特征。In order to process the non-stationary vibration signals such as speed-up or speed-down effectively,a method,which integrates order tracking technique with singular spectrum denoising,is presented to diagnose the early fault signals of gearbox.Firstly,the vibration signals are sampled at constant time increments and the data are resampled at constant angle increments.Therefore,the time-domain transient signals are transformed into angle-domain pseudo-stationary ones.Secondly,the signals in angle-domain are analyzed by using singular spectrum denoising.At last,the denoised signals are analyzed by using order tracking.The experimental results show that the method can effectively diagnose the early gear fault.

关 键 词：阶次跟踪 奇异谱降噪 齿轮 故障诊断 

分 类 号：TH132.4[机械工程—机械制造及自动化]